Summary Analysis of Remote Emission Sensing Data

Author

Naomi Farren, David Carslaw

1 Introduction

The goal of the City Air Remote Emission Sensing CARES project is to reduce the hurdles for the practical application of remote emission sensing and to make it a widespread means for the monitoring and enforcement of vehicle emissions, leading to improvements in air pollution. This document forms part of work package 4, which aims to enable cities and other administrations to set up remote emission sensing (RES) measurements more easily and to quicker analyse the data. Here an open-access software package has been developed, containing a suite of user functions that are designed to answer typical questions on the emission performance of the vehicles measured.

1.1 Aims

The analysis of remote emission sensing (RES) data can be challenging, given the complexity and typical size of the data collected during experimental campaigns. Even within the small community of researchers and practitioners that typically conduct experiments, there is a wide variation in the analysis approaches used and their consistency. With that in mind, this document aims to:

  1. Provide a reliable and automated way of presenting key summary data and plots from RES campaigns.

  2. Adopt ‘modern’ data analysis approaches using R Statistical Software and automated report production using Quarto.

  3. Present common numerical and graphical outputs that help to interpret data from RES campaigns.

The use of R software and Quarto offers many advantages over traditional ways of analysing and presenting data. For example, allowing for detailed data to be presented in a compact way that can be easily filtered by the users, and the use of ‘tabs’ to better structure the output.

openCARES

The analysis software and the underlying code that produced this document are part of a R package called openCARES. The package is available as a GitHub repository and all code is managed under a version control system. The approach means that all changes are recorded and that members of the CARES team can work collaboratively to develop the analysis capabilities over time. Access to the openCARES repository can be found here.

1.2 Data requirements

Example data is used to demonstrate the types of analysis that can be performed using the openCARES package. An evaluation of the vehicle fleet composition and site conditions is provided, in addition to fuel-specific emission factors, grouped by vehicle type, fuel type, emission standard, manufacturer and so on. The effect of ambient temperature and vehicle deterioration on emissions is explored and distance-specific factors are calculated. The example data was collected in 2021 during a city demonstration measurement campaign in Milan, Italy, conducted as part of the CARES project. The data set consists of approximately 35,500 measurements collected using an Emissions Detection and Reporting (EDAR) system, developed by Hager Environmental & Atmospheric Technologies (HEAT). Provided that the data requirements outlined below are met, the analysis can be conducted on other data sets collected using a wide range of RES techniques, such as cross-road remote sensing, plume chasing and point sampling.

Data field specifications

A description of the individual data fields required for analysis is provided in Tables 1-3. The data must be in the correct format and appropriate units prior to running the analysis. The end user can download and edit their own version of the source code if necessary.

Table 1. Required site data fields.
Name Description
site_name Site name
latitude Latitude (dec °)
longitude Longitude (dec °)
altitude Altitude (m)
slope Road slope (°)
amb_temp Ambient temperature (°C)
amb_rhum Relative humidity (%)
Table 2. Required measurement data fields.
Name Description
date_time ISO 8601 e.g. 2021-09-23T16:11:42.10008Z
co_fm Fuel-specific CO emissions (g kg-1)
no_fm Fuel-specific NO emissions (g kg-1)
no2_fm Fuel-specific NO2 emissions (g kg-1)
hc_fm Fuel-specific HC emissions (g kg-1)
ch4_fm Fuel-specific CH4 emissions (g kg-1)
valid_status Valid measurement (TRUE or FALSE)
speed Vehicle speed (km h-1)
vsp_calc Calculated VSP according to U.S. EPA (kW t-1)
Table 3. Required vehicle data fields.
Name Description
fuel_type_1 Primary fuel type e.g. diesel, petrol, CNG, LPG
fuel_type_2 Secondary fuel type for bi-fuel and hybrid e.g. CNG, LPG, electricity
veh_class Vehicle class e.g. passenger car
emission_standard Emission standard e.g. Euro 5
make_domain Manufacturer e.g. Fiat, Nissan, Volkswagen
reg_date_domestic Registration date (yyyy/mm/dd)
veh_category UNECE vehicle category e.g. M1
mileage Odometer ready from technical inspection (km)
Hybrid vehicles

Gaseous exhaust pollutants are emitted when hybrid vehicles rely on the internal combustion engine and the derived fuel-specific emission factors (expressed as grams of pollutant per kilogram of fuel) are representative of this. Average emission factors for an entire journey will be lower, depending on the proportion of time the hybrid vehicle operates in battery mode.

Bi-fuel vehicles

Bi-fuel vehicles, e.g. petrol CNG, petrol LPG, have multi fuel engines that are capable of running on two fuels. The fuels are stored in separate tanks and the engine can run on one fuel at a time. The calculations used to derive fuel-specific emission factors assume that the vehicle is using Compressed Natural Gas (CNG) or Liquid Natural Gas (LPG).

2 Measurement site conditions

This section provides information about the measurement sites used to collect the data, including a summary of the meteorological conditions.

2.1 Site information

Site Name Start date End date Latitude Longitude Altitude (m) Slope
Cilea 2021-09-24 2021-10-07 45.50 9.10 136.00 0.17
Madre Cabrini 2021-09-23 2021-10-08 45.45 9.20 116.00 0.06

2.2 Ambient temperature

Ambient temperature (°C) n
22.02 35.57K

Site name Ambient temperature (°C) n
Cilea 22.77 11.75K
Madre Cabrini 21.45 15.57K

2.3 Relative humidity

Relative humidity (%) n
55.01 35.57K

Site name Relative humidity (%) n
Cilea 49.13 11.75K
Madre Cabrini 59.44 15.57K

2.4 Vehicle dynamics

Speed (km h-1) Acceleration (unit?) Vehicle specific power (kW t-1) n
32.88 0.84 4.57 27.32K

Site name Speed (km h-1) Acceleration (unit?) Vehicle specific power (kW t-1) n
Cilea 42.64 0.42 4.83 11.75K
Madre Cabrini 25.51 1.15 4.38 15.57K

3 Vehicle fleet composition

Note

Measurements are grouped by vehicle class and fuel type. Groups comprising less than 0.5% of the measurements are categorised as ‘Other’.

3.1 Vehicle and fuel type

Tip

Hover over each segment to obtain the number and percentage of measured vehicles. Add or remove groups by clicking on the list in the legend.

3.2 Euro class

3.3 Manufacturers

Note

The manufacturers are assigned to groups e.g. “VWG” includes Audi, Bentley, Lamborghini, Porsche, Seat, Skoda and Volkswagen. The size of each rectangle is proportional to the share of each manufacturer / manufacturer group.

4 Vehicle emissions

4.1 Emissions by Euro class

Note

Emission values are shown when the number of measurements for a particular vehicle type and Euro class is greater than 100.

Tip

NOx represents the sum of NO and NO2. To generate NOx emission factors, NO emission factors are multiplied by 46/30 to generate ’NO as NO2 equivalent` emission factors, which are then added to the NO2 emission factors.

4.2 Emissions by vehicle registration year

Note

Measurements are grouped by vehicle class and fuel type. Groups comprising less than 5% of the measurements are excluded. Emission values are shown when the number of measurements for a particular vehicle registration year within a group is greater than 100.

4.3 Emissions by manufacturer

4.4 Emission summaries

Tip

The pollutant emission factors provided in the emission summary tables and detailed pollutant summary tables are expressed as grams of pollutant per kilogram of fuel. ‘n’ shows the number of measurements in each group.

CO NO NO2 NOx CH4 HC n
10.68 3.77 0.77 4.54 0.16 1.97 27.32K
Site Name CO NO NO2 NOx CH4 HC n
Cilea 14.43 3.91 0.60 4.51 0.25 2.62 11.75K
Madre Cabrini 7.85 3.66 0.90 4.56 0.08 1.48 15.57K
Fuel Type CO NO NO2 NOx CH4 HC n
CNG 11.95 5.18 0.46 5.65 2.79 3.41 145.00
diesel 2.69 6.09 1.60 7.69 0.05 1.45 10.98K
diesel electricity 15.17 0.90 0.30 1.20 0.35 −0.23 2.00
diesel LPG 0.83 11.43 1.84 13.27 0.13 10.18 4.00
LPG 0.44 0.92 −0.14 0.78 0.31 −7.48 1.00
petrol 14.13 1.95 0.17 2.13 0.10 2.17 13.07K
petrol CNG 16.46 4.08 0.38 4.46 3.32 3.79 423.00
petrol electricity 11.28 0.62 0.06 0.68 0.04 1.29 184.00
petrol LPG 32.86 2.90 0.20 3.10 0.27 2.92 1.90K
Fuel Type CO NO NO2 NOx CH4 HC n
Cilea
CNG 13.30 6.43 0.84 7.27 3.26 3.89 61.00
diesel 3.64 6.30 1.26 7.55 0.13 1.82 3.95K
diesel LPG 0.08 4.26 1.48 5.74 0.23 14.65 3.00
petrol 17.59 2.49 0.23 2.73 0.18 2.81 6.25K
petrol CNG 21.60 3.85 0.48 4.33 3.58 4.59 202.00
petrol electricity 18.36 1.72 0.08 1.79 0.10 2.67 40.00
petrol LPG 34.91 3.03 0.28 3.31 0.33 3.91 1.07K
Madre Cabrini
CNG 10.97 4.28 0.19 4.47 2.46 3.06 84.00
diesel 2.15 5.97 1.80 7.77 0.01 1.24 7.03K
diesel electricity 15.17 0.90 0.30 1.20 0.35 −0.23 2.00
diesel LPG 3.05 32.93 2.92 35.85 −0.17 −3.24 1.00
LPG 0.44 0.92 −0.14 0.78 0.31 −7.48 1.00
petrol 10.96 1.46 0.11 1.57 0.03 1.58 6.82K
petrol CNG 11.77 4.29 0.28 4.57 3.08 3.07 221.00
petrol electricity 9.32 0.32 0.05 0.37 0.02 0.90 144.00
petrol LPG 30.21 2.73 0.10 2.83 0.19 1.64 827.00

4.5 Detailed pollutant summaries

5 Ambient temperature effects

Tip

Here we consider the effect of ambient temperature on emissions for vehicle and fuel type groups that make up at least 20% of the total measured vehicle fleet.

6 Vehicle deterioration effects

Note

Vehicle mileage data from annual technical inspection tests may be available. This is considered a good proxy for examining the effect of vehicle deterioration on emissions behaviour since it is a direct measure of the distance a vehicle has driven. Emission measurements associated with mileages less than or equal to 250,000 km are considered here. The effects of vehicle deterioration on emissions above 250,000 km are more uncertain due to the small proportion of measurements available at higher mileages.

7 Distance specific emissions

(Davison et al. 2020)

References

Davison, Jack, Yoann Bernard, Jens Borken-Kleefeld, Naomi J. Farren, Stefan Hausberger, Åke Sjödin, James E. Tate, Adam R. Vaughan, and David C. Carslaw. 2020. “Distance-Based Emission Factors from Vehicle Emission Remote Sensing Measurements.” Science of The Total Environment 739 (October): 139688. https://doi.org/10.1016/j.scitotenv.2020.139688.